{"id":"7f788cff-d350-444a-b8f6-671286b77e06","arxiv_id":"2501.13138","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Simulations across four indoor factory profiles show 5G-TSN can provide bounded delay for latency-sensitive traffic in sparse-clutter, low-base-station settings, with degradation as clutter and base station height increase.","lead":"This paper simulates a 5G network combined with Time-Sensitive Networking in several indoor factory layouts to test whether robot and control traffic can keep low, bounded delays as device count grows. It finds that bounded delay is possible in some layouts, but the result depends heavily on clutter density and base station antenna height.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper reports average end-to-end delay, but the central claim of 'bounded delay' requires worst-case or percentile delay against an explicit latency budget; no such bound or budget is stated.","rationale":"The reader's weakest assumption was simulation fidelity, which is a legitimate concern about whether the simulator reproduces a real 3GPP-standardized 5G-TSN system. The concern raised here is distinct and more direct: even granting perfect simulator fidelity, the reported metric (average delay) does not logically establish the claimed property (bounded delay). TSN's bounded low latency is defined in terms of worst-case scheduling guarantees, not empirical means. The paper's Figs. 5 and 6 show average delays; there is no stated deadline, no percentile, and no worst-case analysis. This is an internal support gap rather than an external validation gap. The reader's verdict of CONDITIONAL remains appropriate, since the paper could be strengthened by adding worst-case/percentile delay against an explicit 5QI budget and by releasing configuration files and seeds for reproducibility. The central claim is plausible and not circular, but it is currently under-supported by the presented data. Therefore I do not recommend changing the verdict, but I would add this missing metric as a condition for acceptance.","tokens_in":8138,"tokens_out":2006,"duration_ms":23425,"concrete_test":"Re-run the simulation for at least the NC traffic stream, extract for each InF profile and UE count the maximum and 99.9th percentile end-to-end delay across packets and across multiple random seeds, and compare these values against an explicit delay budget (e.g., the 5QI DC-GBR delay budget of 10 ms for a typical NC flow). If any configuration exceeds the budget, the bounded-delay claim fails; if all configurations comply, report the margin and the number of replications used.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract and conclusion assert that 5G-TSN can provide 'bounded delay' for latency-sensitive applications in scalable indoor factory settings. However, the evidence presented in Figs. 5 and 6 is only the average end-to-end delay (described in Section IV as 'the time required for a data packet to travel from the source application to the destination application'). TSN's bounded latency is a worst-case property: a deterministic or high-probability upper bound on delay, typically stated as a deadline such as the 5QI delay budget in 3GPP TS 23.501 Table 5.7.4-1. No such deadline is specified for the NC traffic stream, no worst-case or high-percentile delay is reported, and no statistical confidence intervals are given. Consequently, even if the simulation stack perfectly models 3GPP 5G-TSN, the measured averages cannot support a claim of bounded delay. The authors themselves note in Section V that real-world validation is needed, but the more immediate gap is that the presented metric does not match the claimed property: a bound is a guarantee, not an average.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript evaluates the scalability of a simulated 5G-TSN deployment in indoor factory environments. The authors integrate 3GPP TR 38.901 indoor factory (InF) path-loss and LOS probability models into an OMNeT++/Simu5G/5GTQ simulation, with four InF profiles (InF-SL, InF-DL, InF-SH, InF-DH), a single gNB, and 5, 10, 25, or 50 UEs generating three traffic classes (Network Control, Video, Best Effort). Performance is reported as SINR distributions, average end-to-end delay, and HARQ error rates for downlink and uplink. The abstract concludes that 5G-TSN 'has the potential to provide bounded delay for latency-sensitive applications in scalable indoor factory settings.' The paper is framed as, to the authors' knowledge, the first scalability study of 3GPP-standardized 5G-TSN in a wireless manufacturing environment.","tokens_in":8376,"tokens_out":3142,"duration_ms":33197,"significance":"If substantiated, the paper would provide a useful early quantitative baseline for wireless TSN scalability in industrial settings, since it applies standardized InF channel models with a concrete traffic mix and network stack. Its strengths include the use of well-known external frameworks (OMNeT++, Simu5G, 5GTQ) and the explicit reporting of simulation parameters in Table I, which aids reproducibility. However, the actual evidence consists of qualitative trends in average delay and error rates; the central claim of 'bounded delay' is not directly supported by the metrics presented. The contribution is therefore incremental and needs substantial strengthening before the stated conclusions can be accepted.","major_comments":[{"comment":"The central claim that 5G-TSN 'has the potential to provide bounded delay' is not supported by the reported evidence. Figures 5 and 6 show only average end-to-end delay, while TSN bounded latency is a worst-case or high-probability upper-bound property, typically expressed against a deadline such as the 5QI delay budget in 3GPP TS 23.501 Table 5.7.4-1. No target latency deadline is specified for the NC traffic stream, and no maximum, 99th percentile, or tail-delay statistic is reported. Adding worst-case/percentile delay against a concrete budget (e.g., the relevant 5QI delay budget) is essential to justify 'bounded delay' in the abstract and conclusion.","section":"Abstract and §V"},{"comment":"No statistical confidence information is provided: the number of simulation runs, random seeds, confidence intervals, or error bars are absent. The HARQ error rates in Figs. 7 and 8 are ratios of failed to total transmissions, but without run-level variance it is impossible to know whether differences between profiles, such as the claimed lower error rate of InF-DL versus InF-SL at higher UE densities, are significant or simply simulation noise. The paper should report run counts and error bars or confidence intervals for the delay and error-rate results.","section":"§IV, Figs. 5–8"},{"comment":"The manuscript defines three distance regions (d1, d2, d3) and states that delay and HARQ results are analyzed, but Figs. 5–8 appear to use only the d2 constraint (the text says 'a maximum distance (d2) constraint of 170 meters'). The scalability claim across 'indoor factory settings' would require results for d1 and d3, or at least a clear justification for why d2 alone is representative. In addition, the maximum tested density is 50 UEs with a single gNB; this is a limited basis for the strong word 'scalable' in the abstract. Reporting results for the full distance range and more UE counts would make the scalability analysis more convincing.","section":"§IV, distance regions and scalability"},{"comment":"The paper repeatedly describes the simulation as '3GPP compliant' and 'standardised,' but no validation or calibration of the simulator against known 3GPP or field results is provided. Section V itself concedes that 'future efforts could focus on real-world deployment and validation to confirm simulation results.' For a simulation-only study, a sanity check (e.g., comparing simulated path loss or SINR against TR 38.901 reference values, or calibrating the scheduler/HARQ behavior against published 5G results) is needed to support the claim that the model faithfully represents a real 5G-TSN system. Without this, the correctness of the qualitative conclusions rests on the unverified fidelity of the simulator stack.","section":"§IV, simulation fidelity"}],"minor_comments":[{"comment":"The abstract states that 5G offers 'negligible jitter,' but jitter is not measured or reported anywhere in the paper; this claim should be removed or qualified.","section":"Abstract and §IV"},{"comment":"The notation is inconsistent: the text uses 'OMNET++' while the OMNeT++ project uses 'OMNeT++,' and 'iNet' should be 'INET.' Please unify these names.","section":"Throughout"},{"comment":"InF-HH is defined with a LOS probability of 1 in Eq. (8), but InF-HH is not included in the simulations reported in Table I or the figures. The omission should be stated explicitly, or the profile should be included for completeness.","section":"§III, Eqs. (6)–(8)"},{"comment":"The table lists 'Target Bler 0.01' but the standard abbreviation is 'BLER'; also, the choice of numerology index 4 is not explained in the text, even though numerology directly affects subcarrier spacing and symbol duration and hence latency.","section":"Table I"},{"comment":"The HARQ error rate is defined as 'the ratio of failed transmissions to the total number of transmissions,' but it is unclear whether this is per UE, per packet, or aggregated over the whole simulation. Please clarify the denominator and whether retransmissions at different HARQ attempts are counted separately.","section":"§IV, HARQ definition"},{"comment":"The concluding statement that '5G-TSN can reliably support latency-sensitive applications' is stronger than the evidence, since no reliability target (e.g., 99.999% or a specific maximum error probability) is stated. The wording should be aligned with the reported metrics.","section":"§V"},{"comment":"The acknowledgment contains a typo: 'Taighde ireann Research Ireland' should be 'Taighde Éireann Research Ireland.'","section":"Acknowledgments"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the headline: this is a modest but real first step. Nobody else has plugged the TR 38.901 InF profiles into a 5G-TSN simulator and looked at scalability. The paper earns credit for using standard channel models and the OMNeT++/Simu5G/5GTQ stack, and for clearly separating NC, Video, and BE traffic. As a baseline for future work, it is useful.\n\nThe soft spots are mostly about evidence matching the claims. The abstract and conclusion say 5G-TSN can provide 'bounded delay' for latency-sensitive applications. What is actually reported is average end-to-end delay. A TSN delay bound is a worst-case or high-percentile number against an explicit deadline (e.g., a 5QI delay budget). No delay budget, no percentiles, no confidence intervals. So the central claim is not supported by the presented metric. This is not a fatal flaw for a baseline study, but the language should be softened and tail latency added.\n\nAlso missing: the authors define d1/d2/d3 distance regions but all results are for d2 only. So the 'scalability' analysis is over UE count, not distance. Minor: no simulation artifacts or run counts, so reproducibility is outline-level. And the HARQ error rate discussion seems odd — they say InF-SL has moderate errors and InF-DL lower, which sits uneasily with the SINR ranking; either the figure shows something else or the text needs clarification.\n\nCitation pattern is fine: the self-citation [1] is background only, the other references are standard. No circularity concerns.\n\nI would send this to peer review with major-revision expectations. The combination is novel enough, the topic is relevant to industrial wireless, and the flaws are fixable with more careful metrics and a smaller scope claim.","headline":"Useful first simulation baseline combining 3GPP indoor factory channels with 5G-TSN, but the 'bounded delay' claim rests on averages and needs worst-case statistics and artifacts.","tokens_in":8932,"tokens_out":1767,"would_cite":false,"duration_ms":18794,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"5G-TSN can provide bounded delay in simulated indoor factories","keywords":["5G","TSN","Industry 4.0","Wireless TSN","Industrial Networks","Indoor Factory","Smart Factory"],"falsifier":"A real-world trial in an indoor factory with a single base station, 50 moving devices, and the same traffic mix: if network-control packets miss their latency bound in a sparse-clutter, low-base-station layout, the central claim would be falsified.","tokens_in":7932,"feed_emoji":"🏭","tokens_out":7841,"duration_ms":66635,"temperature":0.7,"pith_summary":"This paper argues that integrating fifth-generation mobile networking (5G) with Time-Sensitive Networking (TSN) can keep latency-sensitive industrial traffic within its timing bounds even as the number of wireless devices in an indoor factory grows. The authors simulate a standard-compliant 5G network carrying TSN traffic in four indoor-factory channel profiles that differ in clutter density and base-station height, with up to 50 devices. They find that the sparse-clutter, low-base-station profile gives the highest signal quality and the best delay performance, while dense clutter and high base stations degrade both. The claim is that 5G-TSN is a workable wireless replacement for wired TSN in favorable factory layouts, though the results are simulation-based and not yet experimentally confirmed.","feed_headline":"5G-TSN can provide bounded delay in simulated indoor factories","feed_subtitle":"Sparse clutter and low base stations keep control traffic within latency limits as device counts grow, simulations show.","key_machinery":"The argument runs on three pieces: the indoor-factory channel model that defines signal quality for each layout, using profiles that vary clutter and base-station height; the TSN traffic shaping inside the network, with strict-priority queueing for control traffic, a credit-based shaper for video, and best-effort handling for background; and the 5G bridge architecture that attaches TSN flows to the wireless link. The channel model sets the signal quality, which determines error rates and end-to-end delay, while the traffic shapers decide which packets wait when resources are scarce.","core_discovery":"The central discovery is that a 5G-TSN network can support time-critical network-control traffic with bounded end-to-end delay in an indoor factory, provided the factory has sparse clutter and a low base station. Under the standard indoor-factory channel model, the sparse-clutter, low-base-station profile (InF-SL) sustains the highest signal-to-interference-plus-noise ratio and the lowest latency across device counts from 5 to 50, while the dense-clutter, high-base-station profile (InF-DH) shows the worst performance. The paper shows that at 50 devices, control-traffic delay rises because video transmissions occupy resources, and it proposes strict-priority scheduling combined with preemption to restore bounded delay.","pith_inferences":["If confirmed in hardware, the result implies that factory layout planning should prioritize sparse clutter and low antenna mounting to make wireless TSN viable without extra scheduling complexity.","The study assumes a single base station, so it leaves open how handover between base stations would affect bounded delay in larger factories; a multi-base-station extension would be a natural next test.","The proposed preemptive scheduling could be evaluated in the same simulation before any physical deployment, offering a low-cost check of the mitigation strategy.","Because the results are simulation-only, the quantitative latency values should be treated as indicative rather than guaranteed until measured on real equipment."],"forward_implications":["In the sparse-clutter, low-base-station profile (InF-SL), 5G-TSN maintains high signal quality and low end-to-end delay for network-control traffic as device count rises to 50.","Dense clutter combined with a high base station (InF-DH) degrades signal quality and increases error rates, making such layouts the hardest for wireless TSN.","At 50 UEs, video traffic competes with control traffic for radio resources and raises control-packet delay; the paper suggests preemptive scheduling to fix this.","Uplink and downlink behave similarly, with low base-station placement giving better reliability in both directions."],"supporting_citations":[{"why":"Supplies the integrated 5G-TSN bridge architecture that the simulation models.","marker":"[2]"},{"why":"Provides the indoor-factory channel model and InF profile parameters used in the simulation.","marker":"[3]"},{"why":"Defines the automated guided vehicle use case and traffic requirements the scenarios follow.","marker":"[4]"},{"why":"Gives the discrete-event simulation environment in which the network model runs.","marker":"[5]"},{"why":"Supplies the 5G New Radio simulation model for the radio access network.","marker":"[7]"},{"why":"Provides the QoS-aware 5G-TSN simulation framework that bridges TSN and 5G in the model.","marker":"[8]"},{"why":"Defines the standardized 5QI-to-QoS mapping used to prioritize traffic classes.","marker":"[9]"}],"fun_headline_variants":["5G-TSN scales with bounded delay in simulated factory","Sparse clutter keeps 5G-TSN latency in check","Indoor factory 5G-TSN: bounded delay for control traffic","Simulation: 5G-TSN suits sparse indoor factory settings","5G-TSN bounded delay holds across device counts in factory"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the simulation stack accurately reproduces the behavior of a real standard-compliant 5G-TSN system, including radio scheduling and error recovery; the paper's own conclusion notes that real-world deployment and validation are still needed.","fun_headline_variants_meta":{"raw":{"variants":["5G-TSN scales with bounded delay in simulated factory","Sparse clutter keeps 5G-TSN latency in check","Indoor factory 5G-TSN: bounded delay for control traffic","Simulation: 5G-TSN suits sparse indoor factory settings","5G-TSN bounded delay holds across device counts in factory"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000687,"raw_usage":{"total_tokens":3073,"prompt_tokens":861,"completion_tokens":2212,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":477,"completion_tokens_details":{"reasoning_tokens":2124}},"tokens_in":477,"tokens_out":2212,"duration_ms":15763,"temperature":1.0,"reasoning_tokens":2124,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T16:43:30.898959+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A real-world trial in an indoor factory with a single base station, 50 moving devices, and the same traffic mix: if network-control packets miss their latency bound in a sparse-clutter, low-base-station layout, the central claim would be falsified.","supporting_citations":[{"cited_title":"A review of recent advances in automated guided vehicle technologies: Integration challenges and research areas for 5G-Based smart manufacturing applications,","cited_arxiv_id":null,"evidence_quote":"Defines the automated guided vehicle use case and traffic requirements the scenarios follow."},{"cited_title":"Performance of integrated 3GPP 5G and IEEE TSN networks,","cited_arxiv_id":null,"evidence_quote":"Supplies the integrated 5G-TSN bridge architecture that the simulation models."},{"cited_title":"Study on channel model for frequencies from 0.5 to 100 GHz,","cited_arxiv_id":null,"evidence_quote":"Provides the indoor-factory channel model and InF profile parameters used in the simulation."},{"cited_title":"An overview of the OMNeT++ simulation environment,","cited_arxiv_id":null,"evidence_quote":"Gives the discrete-event simulation environment in which the network model runs."},{"cited_title":"Simu5G: Simulator for 5G new radio networks,","cited_arxiv_id":null,"evidence_quote":"Supplies the 5G New Radio simulation model for the radio access network."},{"cited_title":"5GTQ: QoS- Aware 5G-TSN simulation framework,","cited_arxiv_id":null,"evidence_quote":"Provides the QoS-aware 5G-TSN simulation framework that bridges TSN and 5G in the model."},{"cited_title":"Technical Specification Group Services and System Aspects; System Architecture for the 5G System (3GPP TS 23.501 version 18.7.0 Release 18),","cited_arxiv_id":null,"evidence_quote":"Defines the standardized 5QI-to-QoS mapping used to prioritize traffic classes."}],"review_version":1}